Philips CEO: AI Transforms Healthcare, Trust is Key Barrier
Artificial intelligence is quietly revolutionizing the efficiency and potential of U.S. healthcare, even amidst significant shifts in government policy and spending. At the forefront of this transformation is Philips, the venerable electronics manufacturer that has evolved into a leading medtech provider. Jeff DiLullo, CEO of Philips North America, offers insights into how AI is already delivering tangible improvements in patient care, from enhancing radiology scans to accelerating cancer diagnoses, and underscores the imperative for industry leaders to adapt their operational models to meet the current moment.
While AI dominates contemporary discourse, its practical implementation varies across sectors. In medtech, DiLullo observes a nuanced landscape: certain AI applications are remarkably mature, having received FDA clearance and proving safe for clinical use. Other areas remain experimental. However, a pervasive challenge across the board is the nascent level of trust in AI, which currently stands as the most significant barrier to its widespread adoption.
This trust deficit is starkly highlighted by the 2025 Future Health Index, which reveals a significant disparity: approximately 60 to 65 percent of clinicians express trust in AI, yet only about one-third of patients, particularly older demographics, share this confidence. Bridging this gap is critical. DiLullo notes that younger generations, being “digitally fluid,” inherently adapt more easily to AI models. For older patients, the key lies with the healthcare practitioners themselves. If doctors and nurses believe in AI’s credibility and utilize it to augment—rather than replace—their analytical and diagnostic capabilities, patient trust is expected to rise. Philips’s role, according to DiLullo, is to provide robust, FDA-cleared AI diagnostic tools, confident that if clinicians perceive value in increased patient interaction time and reduced stress, adoption will accelerate exponentially.
The practical application of AI in healthcare is already evident in fields like radiology, where early diagnosis is paramount for optimal outcomes. Despite technological advancements, patients often face lengthy waits for scans, sometimes exceeding a month. Philips is addressing this with AI-powered MRI systems featuring “Smart Speed” technology. This innovation significantly reduces scan times—a procedure that once took 45 minutes can now be completed in just 20. The AI within these systems doesn’t merely fill in gaps; it intelligently filters out noise, yielding clearer, more precise images in less time. For radiologists, this translates to increased productivity, enabling them to process more studies daily (e.g., from 12-15 to 20), thereby improving patient throughput, enhancing hospital reimbursement, and ultimately, elevating patient care.
Beyond image acquisition, AI streamlines the diagnostic workflow. It can pinpoint anomalies in digital images, guiding radiologists directly to areas requiring closer examination. The advent of digital pathology further transforms the process, allowing for rapid cancer diagnoses within hours—a significant leap from traditional methods. This digital workflow and orchestration represent a fundamental shift in how healthcare is delivered.
While the potential for “AI hallucinations” found in some generative AI models raises valid concerns, DiLullo stresses that current, proven AI applications in healthcare are already delivering substantial benefits. Features like Smart Speed, compressed diagnostic timelines, and on-demand “tumor board” meetings are available today, albeit not yet adopted at their full potential across all health systems. He urges institutions to implement these existing solutions. Regarding more experimental generative AI models, while caution, robust controls, and governance are essential, refraining from experimentation is not an option. Leading institutions, including MGB, Stanford, and Mount Sinai, are actively engaging with population health data to train AI models for both specific and broad use cases.
DiLullo emphasizes that healthcare doesn’t need to wait for a “silver bullet” solution to achieve immortality or solve every health problem. Instead, the immediate focus should be on optimizing the existing system. Drawing an analogy to learning to drive, he suggests mastering local roads before attempting the Autobahn. A substantial 80 percent of AI’s potential for driving productivity at scale can be realized today through mature AI and virtual capabilities. This immediate, tangible impact represents the next significant opportunity for healthcare, driven by the profound and ever-growing need for more efficient, effective patient care.